292 Etltel McLenvan : 



The fungus in the seed-coat becomes active on gerniination, 

 infects the seedling, and produces the felt on the root; indeed the 

 association is so close that the symbiosis is an obligative one, the 

 seedlings not developing unless infected. 



This is a very striking instance of symbiotic association, and, as 

 Kayner points out, the only other plant for wliich a like distribu- 

 tion of the fungus has l>een described is Loliiim temitltntum. 

 Since I have demonstrated the intra-cellular nature of the Lolium 

 fungus it falls even more in line with the Calluna tyjie. Although 

 it does not form a mycorrliiza, its use to the plant is certainly 

 demonstrated at the fruiting period. It is possible, therefore, to 

 regard it as a case parallel to the Calluna type, the similarity to 

 the Ustilagine mode of growth being accidental, and not of any 

 real importance in helping us to classify it and to grasp its affini- 

 ties. Since the Calluna fungus has yielded to artificial culture, and 

 can be classed definitely as a Plioma, every hope can l>e entertained 

 for success in this direction as regards the fungus of the Loliums. 



The occasional penetrations of the endosperm, etc., already de- 

 scribed, do not, I tliink. point to vestigial traces of a former 

 parasitic habit. Even when present they do not evince any haj-ni- 

 ful results in the grain. They are probably to be explained as a 

 luxuriant development of the fungus, resulting perhaps from good 

 growth conditions. Ihe endosperm has proved unable to cope with 

 the large food supply represented by the fungus, and conse- 

 quently has failed to transform it all into the usual storage form — 

 starch and aleurone grains, so that some of the hyphae which had 

 penetrated the tissues of the embryo, and would normally have been 

 absorbed, remained intact. Any food-material they contained 

 would be yielded up on germination, just as the food-material of 

 the endosperm is changed into solu])le forms, and translocated to the 

 ■seat of growth. 



Although I have never examined a grain of Lolium ptren?ie 

 •either mature or in an embryonic condition without finding tlifo 

 fungus present (sometimes in minute amounts), it isi quite probable 

 that such occur. 



It is conceivable that the hyphae, when growing in the young 

 inflorescence may miss a carpel and in tliat case the ovule would 

 not become infected. This would probably not prevent the forma- 

 tion of a fruit. Tlie grain so formed, however, would not be so 

 well equipped in its struggle for existence as its fungal-containing 

 neighbour, and eventually Avould tend to die out. It is, therefore, 



